Annotation of src/usr.bin/sort/sort.c, Revision 1.17
1.17 ! millert 1: /* $OpenBSD: sort.c,v 1.16 2001/02/04 21:27:01 ericj Exp $ */
1.1 millert 2:
3: /*-
4: * Copyright (c) 1993
5: * The Regents of the University of California. All rights reserved.
6: *
7: * This code is derived from software contributed to Berkeley by
8: * Peter McIlroy.
9: *
10: * Redistribution and use in source and binary forms, with or without
11: * modification, are permitted provided that the following conditions
12: * are met:
13: * 1. Redistributions of source code must retain the above copyright
14: * notice, this list of conditions and the following disclaimer.
15: * 2. Redistributions in binary form must reproduce the above copyright
16: * notice, this list of conditions and the following disclaimer in the
17: * documentation and/or other materials provided with the distribution.
18: * 3. All advertising materials mentioning features or use of this software
19: * must display the following acknowledgement:
20: * This product includes software developed by the University of
21: * California, Berkeley and its contributors.
22: * 4. Neither the name of the University nor the names of its contributors
23: * may be used to endorse or promote products derived from this software
24: * without specific prior written permission.
25: *
26: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36: * SUCH DAMAGE.
37: */
38:
39: #ifndef lint
40: static char copyright[] =
41: "@(#) Copyright (c) 1993\n\
42: The Regents of the University of California. All rights reserved.\n";
43: #endif /* not lint */
44:
45: #ifndef lint
46: #if 0
47: static char sccsid[] = "@(#)sort.c 8.1 (Berkeley) 6/6/93";
48: #else
1.17 ! millert 49: static char rcsid[] = "$OpenBSD: sort.c,v 1.16 2001/02/04 21:27:01 ericj Exp $";
1.1 millert 50: #endif
51: #endif /* not lint */
52:
1.4 millert 53: /*
54: * Sort sorts a file using an optional user-defined key.
1.1 millert 55: * Sort uses radix sort for internal sorting, and allows
56: * a choice of merge sort and radix sort for external sorting.
57: */
58:
59: #include "sort.h"
60: #include "fsort.h"
61: #include "pathnames.h"
62:
1.13 espie 63: #include <sys/types.h>
64: #include <sys/stat.h>
1.16 ericj 65: #include <locale.h>
1.1 millert 66: #include <paths.h>
67: #include <signal.h>
68: #include <stdlib.h>
69: #include <string.h>
70: #include <unistd.h>
1.10 mickey 71: #include <err.h>
1.1 millert 72:
73: int REC_D = '\n';
74: u_char d_mask[NBINS]; /* flags for rec_d, field_d, <blank> */
1.4 millert 75:
1.1 millert 76: /*
77: * weight tables. Gweights is one of ascii, Rascii..
78: * modified to weight rec_d = 0 (or 255)
79: */
80: extern u_char gweights[NBINS];
81: u_char ascii[NBINS], Rascii[NBINS], RFtable[NBINS], Ftable[NBINS];
1.4 millert 82:
1.1 millert 83: /*
84: * masks of ignored characters. Alltable is 256 ones
85: */
86: u_char dtable[NBINS], itable[NBINS], alltable[NBINS];
87: int SINGL_FLD = 0, SEP_FLAG = 0, UNIQUE = 0;
1.12 millert 88: struct coldesc *clist;
1.1 millert 89: int ncols = 0;
1.12 millert 90: int ND = 10; /* limit on number of -k options. */
1.1 millert 91:
92: char devstdin[] = _PATH_STDIN;
93: char toutpath[_POSIX_PATH_MAX];
94: char *tmpdir = _PATH_VARTMP;
95:
96: static void cleanup __P((void));
97: static void onsig __P((int));
98: static void usage __P((char *));
99:
1.12 millert 100: #define CHECK_NFIELDS \
101: if (++nfields == ND) { \
102: ND += 10; \
103: if ((p = realloc(fldtab, ND)) == NULL) \
104: errx(2, "cannot allocate memory"); \
105: ftpos = p + (ftpos - fldtab); \
106: fldtab = p; \
107: }
108:
1.1 millert 109: int
110: main(argc, argv)
111: int argc;
112: char *argv[];
113: {
114: int (*get)();
115: int ch, i, stdinflag = 0, tmp = 0;
1.12 millert 116: char nfields = 0, cflag = 0, mflag = 0;
1.1 millert 117: char *outfile, *outpath = 0;
1.12 millert 118: struct field *fldtab, *ftpos;
1.1 millert 119: union f_handle filelist;
120: FILE *outfp = NULL;
1.12 millert 121: void *p;
1.4 millert 122:
1.16 ericj 123: setlocale(LC_ALL, "");
124:
1.12 millert 125: if ((clist = calloc((ND+1)*2, sizeof(struct coldesc))) == NULL ||
126: (ftpos = fldtab = calloc(ND+2, sizeof(struct field))) == NULL)
127: errx(2, "cannot allocate memory");
1.1 millert 128: memset(d_mask, 0, NBINS);
129: d_mask[REC_D = '\n'] = REC_D_F;
130: SINGL_FLD = SEP_FLAG = 0;
131: d_mask['\t'] = d_mask[' '] = BLANK | FLD_D;
132: fixit(&argc, argv);
133: if (!issetugid() && (outfile = getenv("TMPDIR")))
134: tmpdir = outfile;
135: while ((ch = getopt(argc, argv, "bcdfik:mHno:rR:t:T:uy:")) != -1) {
1.4 millert 136: switch (ch) {
1.1 millert 137: case 'b': fldtab->flags |= BI | BT;
138: break;
139: case 'd':
1.4 millert 140: case 'f':
1.1 millert 141: case 'i':
1.5 millert 142: case 'n':
1.1 millert 143: case 'r': tmp |= optval(ch, 0);
144: if (tmp & R && tmp & F)
145: fldtab->weights = RFtable;
146: else if (tmp & F)
147: fldtab->weights = Ftable;
1.4 millert 148: else if (tmp & R)
1.1 millert 149: fldtab->weights = Rascii;
150: fldtab->flags |= tmp;
151: break;
152: case 'o':
153: outpath = optarg;
154: break;
155: case 'k':
1.12 millert 156: CHECK_NFIELDS;
1.5 millert 157: setfield(optarg, ++ftpos, fldtab->flags);
1.1 millert 158: break;
159: case 't':
160: if (SEP_FLAG)
161: usage("multiple field delimiters");
162: SEP_FLAG = 1;
163: d_mask[' '] &= ~FLD_D;
164: d_mask['\t'] &= ~FLD_D;
165: d_mask[(int)*optarg] |= FLD_D;
166: if (d_mask[(int)*optarg] & REC_D_F)
167: err(2, "record/field delimiter clash");
168: break;
169: case 'R':
170: if (REC_D != '\n')
171: usage("multiple record delimiters");
172: if ('\n' == (REC_D = *optarg))
173: break;
174: d_mask['\n'] = d_mask[' '];
175: d_mask[REC_D] = REC_D_F;
176: break;
177: case 'T':
178: tmpdir = optarg;
179: break;
180: case 'u':
181: UNIQUE = 1;
182: break;
183: case 'c':
184: cflag = 1;
185: break;
186: case 'm':
187: mflag = 1;
188: break;
189: case 'H':
190: PANIC = 0;
191: break;
192: case 'y':
193: /* accept -y for backwards compat. */
194: break;
195: case '?':
1.8 deraadt 196: default:
197: usage(NULL);
1.1 millert 198: }
199: }
1.4 millert 200:
1.1 millert 201: if (cflag && argc > optind+1)
202: errx(2, "too many input files for -c option");
1.4 millert 203:
1.1 millert 204: if (argc - 2 > optind && !strcmp(argv[argc-2], "-o")) {
205: outpath = argv[argc-1];
206: argc -= 2;
207: }
1.4 millert 208:
1.1 millert 209: if (mflag && argc - optind > (MAXFCT - (16+1))*16)
210: errx(2, "too many input files for -m option");
1.4 millert 211:
1.1 millert 212: for (i = optind; i < argc; i++) {
213: /* allow one occurrence of /dev/stdin */
214: if (!strcmp(argv[i], "-") || !strcmp(argv[i], devstdin)) {
215: if (stdinflag)
216: warnx("ignoring extra \"%s\" in file list",
217: argv[i]);
218: else {
219: stdinflag = 1;
220: argv[i] = devstdin;
221: }
222: } else if ((ch = access(argv[i], R_OK)))
1.14 millert 223: err(2, "%s", argv[i]);
1.1 millert 224: }
1.4 millert 225:
1.5 millert 226: if (!(fldtab->flags & (I|D|N) || fldtab[1].icol.num)) {
1.1 millert 227: SINGL_FLD = 1;
228: fldtab[0].icol.num = 1;
229: } else {
230: if (!fldtab[1].icol.num) {
1.12 millert 231: CHECK_NFIELDS;
1.1 millert 232: fldtab[0].flags &= ~(BI|BT);
233: setfield("1", ++ftpos, fldtab->flags);
234: }
235: fldreset(fldtab);
236: fldtab[0].flags &= ~F;
237: }
238: settables(fldtab[0].flags);
239: num_init();
240: fldtab->weights = gweights;
1.4 millert 241:
1.3 deraadt 242: if (optind == argc) {
243: static char *names[2];
244:
245: names[0] = devstdin;
246: names[1] = NULL;
247: filelist.names = names;
248: optind--;
249: } else
250: filelist.names = argv+optind;
1.4 millert 251:
1.1 millert 252: if (SINGL_FLD)
253: get = makeline;
254: else
255: get = makekey;
1.4 millert 256:
1.1 millert 257: if (cflag) {
258: order(filelist, get, fldtab);
259: /* NOT REACHED */
260: }
1.4 millert 261:
1.1 millert 262: if (!outpath) {
263: (void)snprintf(toutpath,
264: sizeof(toutpath), "%sstdout", _PATH_DEV);
265: outfile = outpath = toutpath;
266: } else if (!(ch = access(outpath, 0)) &&
267: strncmp(_PATH_DEV, outpath, 5)) {
1.17 ! millert 268: struct sigaction act;
1.1 millert 269: int sigtable[] = {SIGHUP, SIGINT, SIGPIPE, SIGXCPU, SIGXFSZ,
270: SIGVTALRM, SIGPROF, 0};
271: int outfd;
1.13 espie 272: mode_t um;
1.4 millert 273:
1.1 millert 274: errno = 0;
1.4 millert 275:
1.1 millert 276: if (access(outpath, W_OK))
1.14 millert 277: err(2, "%s", outpath);
1.4 millert 278: (void)snprintf(toutpath, sizeof(toutpath), "%sXXXXXXXXXX",
279: outpath);
1.13 espie 280: /* use default umask to try and avoid one syscall */
281: um = umask(S_IWGRP|S_IWOTH);
282: if (um != S_IWGRP|S_IWOTH)
283: (void)umask(um);
284: if ((outfd = mkstemp(toutpath)) == -1 ||
285: fchmod(outfd, DEFFILEMODE & ~um) == -1 ||
1.1 millert 286: (outfp = fdopen(outfd, "w")) == 0)
1.14 millert 287: err(2, "%s", toutpath);
1.1 millert 288: outfile = toutpath;
1.4 millert 289:
1.1 millert 290: (void)atexit(cleanup);
1.17 ! millert 291: sigfillset(&act.sa_mask);
! 292: act.sa_flags = SA_RESTART;
! 293: act.sa_handler = onsig;
1.1 millert 294: for (i = 0; sigtable[i]; ++i) /* always unlink toutpath */
295: sigaction(sigtable[i], &act, 0);
296: } else
297: outfile = outpath;
298: if (outfp == NULL && (outfp = fopen(outfile, "w")) == NULL)
1.14 millert 299: err(2, "%s", outfile);
1.1 millert 300: if (mflag)
301: fmerge(-1, filelist, argc-optind, get, outfp, putline, fldtab);
302: else
303: fsort(-1, 0, filelist, argc-optind, outfp, fldtab);
304: if (outfile != outpath) {
305: if (access(outfile, 0))
1.14 millert 306: err(2, "%s", outfile);
1.1 millert 307: (void)unlink(outpath);
308: if (link(outfile, outpath))
309: err(2, "cannot link %s: output left in %s",
310: outpath, outfile);
311: (void)unlink(outfile);
312: }
313: exit(0);
314: }
315:
316: static void
317: onsig(s)
318: int s;
319: {
1.4 millert 320:
1.1 millert 321: cleanup();
1.15 deraadt 322: _exit(2); /* return 2 on error/interrupt */
1.1 millert 323: }
324:
325: static void
326: cleanup()
327: {
1.4 millert 328:
1.1 millert 329: if (toutpath[0])
330: (void)unlink(toutpath);
331: }
332:
333: static void
334: usage(msg)
335: char *msg;
336: {
1.8 deraadt 337: extern char *__progname;
1.4 millert 338:
1.16 ericj 339: if (msg != NULL)
1.14 millert 340: warnx("%s", msg);
1.11 deraadt 341: (void)fprintf(stderr, "usage: %s [-T dir] [-o output] [-cmubdfinrH] "
1.9 deraadt 342: "[-t char] [-R char] [-k keydef] ... [files]\n", __progname);
1.1 millert 343: exit(2);
344: }